Server migration

As my free AWS tier was about to finish, I needed to find a new home for this page!
I looked for any cheap-but-not-dodgy-VPS and ended up subscribing to the cheapest Hetzner plan. Should still be vastly enough.
Signup and server creation was really fast, even though I got flagged as suspicious and had to upload an ID, the whole process took under 10 minutes.
All the apps *should* be working now, drop me a word if there's any bug.

Todo/Pomodoro webapp

As I am going to start a new gig working on a Python backend for a web application, I set out to build a small thing on my own to get a feel for the technology.

  • Python 3.5
  • Bottle Framework
  • SQLite (demo: in memory within Docker)
  • Plain HTML generated with Jinja2
Here is the result: On Github
And the live demo. It is currently deployed in a Docker on this very machine. There's no user registration or login, it might get very ugly and/or get destroyed!

Hikes to be done - Trekking guide

This will probably not be of much interest to many, but for the few who are preparing for the "Prébatoire Accompagnateur Moyenne Montagne", here's something of interest.
One of the pre-requisite for the exam is to hike through ten routes from this list. Thanks to the previous leaflet project, I've been able to put all the routes on a map quickly. I found it to be a great way to discover and plan these hikes.
Direct link

  • Do not load every GPX trace as a different request => Done, one GPX file is now enough. Thanks awk!
  • Reduce routes size using some well known line simplification algorithm (or use existing lib to work on GPX files?)
  • Put generating code on GitHub
  • Show a table with height gain and route length. Allow filtering.

Far from the madding crowd

Ever wondered how far away you can get from the paved roads that get more and more numerous around us?
Well, I did. And I wanted an answer! What's next?
Easy, let's build an app for that! Using data from OpenStreetMaps, we can filter road informations only, load all that into a PostgreSQL database with the right set of GIS (geographical information system) plugins, write a simple backend application in NodeJS, an even simpler frontend that displays it... And here we go!
All the code can be seen here
And the live application here

On the easiness of existing on the web: this page's inception

On a rainy afternoon in Annecy, I remembered how I missed the good old personal webpages from the 2000's, when I discovered the Internet with our family's first 56kbps modem.

Instead of lamenting, why not going with my own?
Turns out it's surprinsingly easy to do so today, with a very limited knowledge of any frontend practices!
Technologies and infrastructure that I used:

  • AWS free tier, with a t2.micro instance
  • Ubuntu LTS, auto-upgrades enabled
  • GoDaddy for namespace registar. 1€ for the first year, not much to loose!
  • Let's Encrypt ACME certbot
  • Plain HTML, CSS from Yahoo's PureCSS
What I ended up with is a simple, fast and not so ugly[dubious][citation needed] webpage to host my stuff. Great!
Let's see how it goes, if my nginx gets pwned by kiddies, if there's any hibben maintenance burden... And last but not least, will I actually use the page?